Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models

Sanne A.E. Peters, Michiel L. Bots, Hester M. den Ruijter, Mike K. Palmer, Diederick E. Grobbee, John R. Crouse, Daniel H. O'Leary, Gregory W. Evans, Joel S. Raichlen, Karel G.M. Moons, Hendrik Koffijberg

Research output: Contribution to journalArticleAcademicpeer-review

49 Citations (Scopus)

Abstract

Objective: To assess the added value of multiple imputation (MI) of missing repeated outcomes measures in longitudinal data sets analyzed with linear mixed-effects (LME) models.

Study Design and Setting: Data were used from a trial on the effects of Rosuvastatin on rate of change in carotid intima-media thickness (CIMT). The reference treatment effect was derived from a complete data set. Scenarios and proportions of missing values in CIMT measurements were applied and LME analyses were used before and after MI. The added value of MI, in terms of bias and precision, was assessed using the mean-squared error (MSE) of the treatment effects and coverage of the 95% confidence interval.

Results: The reference treatment effect was -0.0177 mm/y. The MSEs for LME analysis without and with MI were similar in scenarios with up to 40% missing values. Coverage was large in all scenarios and was similar for LME with and without MI.

Conclusion: Our study empirically shows that MI of missing end point data before LME analyses does not increase precision in the estimated rate of change in the end point. Hence, MI had no added value in this setting and standard LME modeling remains the method of choice.

Original languageEnglish
Pages (from-to)686-695
Number of pages10
JournalJournal of clinical epidemiology
Volume65
Issue number6
DOIs
Publication statusPublished - Jun 2012

Fingerprint

Carotid Intima-Media Thickness
Outcome Assessment (Health Care)
Confidence Intervals
Datasets
Rosuvastatin Calcium

Keywords

  • Carotid intima-media thickness
  • Clinical trials
  • Linear mixed-effects model
  • Methodology
  • Missing repeated outcome measurements
  • Multiple imputation

Cite this

Peters, Sanne A.E. ; Bots, Michiel L. ; den Ruijter, Hester M. ; Palmer, Mike K. ; Grobbee, Diederick E. ; Crouse, John R. ; O'Leary, Daniel H. ; Evans, Gregory W. ; Raichlen, Joel S. ; Moons, Karel G.M. ; Koffijberg, Hendrik. / Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models. In: Journal of clinical epidemiology. 2012 ; Vol. 65, No. 6. pp. 686-695.
@article{d63900960d6543ff8a85bbb3bf5089f9,
title = "Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models",
abstract = "Objective: To assess the added value of multiple imputation (MI) of missing repeated outcomes measures in longitudinal data sets analyzed with linear mixed-effects (LME) models.Study Design and Setting: Data were used from a trial on the effects of Rosuvastatin on rate of change in carotid intima-media thickness (CIMT). The reference treatment effect was derived from a complete data set. Scenarios and proportions of missing values in CIMT measurements were applied and LME analyses were used before and after MI. The added value of MI, in terms of bias and precision, was assessed using the mean-squared error (MSE) of the treatment effects and coverage of the 95{\%} confidence interval.Results: The reference treatment effect was -0.0177 mm/y. The MSEs for LME analysis without and with MI were similar in scenarios with up to 40{\%} missing values. Coverage was large in all scenarios and was similar for LME with and without MI.Conclusion: Our study empirically shows that MI of missing end point data before LME analyses does not increase precision in the estimated rate of change in the end point. Hence, MI had no added value in this setting and standard LME modeling remains the method of choice.",
keywords = "Carotid intima-media thickness, Clinical trials, Linear mixed-effects model, Methodology, Missing repeated outcome measurements, Multiple imputation",
author = "Peters, {Sanne A.E.} and Bots, {Michiel L.} and {den Ruijter}, {Hester M.} and Palmer, {Mike K.} and Grobbee, {Diederick E.} and Crouse, {John R.} and O'Leary, {Daniel H.} and Evans, {Gregory W.} and Raichlen, {Joel S.} and Moons, {Karel G.M.} and Hendrik Koffijberg",
year = "2012",
month = "6",
doi = "10.1016/j.jclinepi.2011.11.012",
language = "English",
volume = "65",
pages = "686--695",
journal = "Journal of clinical epidemiology",
issn = "0895-4356",
publisher = "Elsevier",
number = "6",

}

Peters, SAE, Bots, ML, den Ruijter, HM, Palmer, MK, Grobbee, DE, Crouse, JR, O'Leary, DH, Evans, GW, Raichlen, JS, Moons, KGM & Koffijberg, H 2012, 'Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models' Journal of clinical epidemiology, vol. 65, no. 6, pp. 686-695. https://doi.org/10.1016/j.jclinepi.2011.11.012

Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models. / Peters, Sanne A.E.; Bots, Michiel L.; den Ruijter, Hester M.; Palmer, Mike K.; Grobbee, Diederick E.; Crouse, John R.; O'Leary, Daniel H.; Evans, Gregory W.; Raichlen, Joel S.; Moons, Karel G.M.; Koffijberg, Hendrik.

In: Journal of clinical epidemiology, Vol. 65, No. 6, 06.2012, p. 686-695.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Multiple imputation of missing repeated outcome measurements did not add to linear mixed-effects models

AU - Peters, Sanne A.E.

AU - Bots, Michiel L.

AU - den Ruijter, Hester M.

AU - Palmer, Mike K.

AU - Grobbee, Diederick E.

AU - Crouse, John R.

AU - O'Leary, Daniel H.

AU - Evans, Gregory W.

AU - Raichlen, Joel S.

AU - Moons, Karel G.M.

AU - Koffijberg, Hendrik

PY - 2012/6

Y1 - 2012/6

N2 - Objective: To assess the added value of multiple imputation (MI) of missing repeated outcomes measures in longitudinal data sets analyzed with linear mixed-effects (LME) models.Study Design and Setting: Data were used from a trial on the effects of Rosuvastatin on rate of change in carotid intima-media thickness (CIMT). The reference treatment effect was derived from a complete data set. Scenarios and proportions of missing values in CIMT measurements were applied and LME analyses were used before and after MI. The added value of MI, in terms of bias and precision, was assessed using the mean-squared error (MSE) of the treatment effects and coverage of the 95% confidence interval.Results: The reference treatment effect was -0.0177 mm/y. The MSEs for LME analysis without and with MI were similar in scenarios with up to 40% missing values. Coverage was large in all scenarios and was similar for LME with and without MI.Conclusion: Our study empirically shows that MI of missing end point data before LME analyses does not increase precision in the estimated rate of change in the end point. Hence, MI had no added value in this setting and standard LME modeling remains the method of choice.

AB - Objective: To assess the added value of multiple imputation (MI) of missing repeated outcomes measures in longitudinal data sets analyzed with linear mixed-effects (LME) models.Study Design and Setting: Data were used from a trial on the effects of Rosuvastatin on rate of change in carotid intima-media thickness (CIMT). The reference treatment effect was derived from a complete data set. Scenarios and proportions of missing values in CIMT measurements were applied and LME analyses were used before and after MI. The added value of MI, in terms of bias and precision, was assessed using the mean-squared error (MSE) of the treatment effects and coverage of the 95% confidence interval.Results: The reference treatment effect was -0.0177 mm/y. The MSEs for LME analysis without and with MI were similar in scenarios with up to 40% missing values. Coverage was large in all scenarios and was similar for LME with and without MI.Conclusion: Our study empirically shows that MI of missing end point data before LME analyses does not increase precision in the estimated rate of change in the end point. Hence, MI had no added value in this setting and standard LME modeling remains the method of choice.

KW - Carotid intima-media thickness

KW - Clinical trials

KW - Linear mixed-effects model

KW - Methodology

KW - Missing repeated outcome measurements

KW - Multiple imputation

UR - http://www.scopus.com/inward/record.url?scp=84860236915&partnerID=8YFLogxK

U2 - 10.1016/j.jclinepi.2011.11.012

DO - 10.1016/j.jclinepi.2011.11.012

M3 - Article

VL - 65

SP - 686

EP - 695

JO - Journal of clinical epidemiology

JF - Journal of clinical epidemiology

SN - 0895-4356

IS - 6

ER -